Grok 4.7 vs Kimi K3 256K
Grok 4.7 leads overall, C- to D. The two cost about the same per task. By suite, Kimi K3 256K takes real-world issues (C- to D) and Grok 4.7 takes spec planning (C to D+). Take Kimi K3 256K anyway when you want the side that is faster on real-world issues and ahead on real-world issues.Measured
Each agent graded on the same private suites, running headless through its own tools. Advantages are oriented so a win is a win, whichever way the metric runs.
Verdict: Kimi K3 256K vs Grok 4.7
Grok 4.7 leads overall, C- to D. By suite: Kimi K3 256K takes real-world issues, C- to D (57% vs 47% capability); Grok 4.7 takes spec planning, C to D+ (68% vs 53% capability); Grok 4.7 takes vibe coding, C to F (64% vs 24% capability). Grok 4.7 used 12% of SuperGrok Plus's weekly limit per pass, $0.145 per task at the listed price. Kimi K3 256K used 34% of Kimi Allegretto's weekly limit per pass, $0.161 per task at the listed price. Per task the two cost about the same ($0.145 vs $0.161). Kimi K3 256K is faster on real-world issues: a median task takes 4m 5s against 6m 57s. Both finished every task on it. Reasons to pick Kimi K3 256K anyway: it is faster on real-world issues (4m 5s vs 6m 57s median) and it leads on real-world issues (C- to D).
Changed since W38: Grok 4.7 has one graded week under this version (W39); Kimi K3 256K: real-world issues usage 22% to 23% of the weekly limit, spec planning D to D+, spec planning usage 9% to 11% of the weekly limit, vibe coding D+ to C-, vibe coding usage 2.6% to 3% of the weekly limit.
| Dimension | Grok 4.7 | Kimi K3 256K | Advantage |
|---|---|---|---|
| Capability | 47% | 57% | +10% (Kimi K3 256K leads) |
| Median task time | 6m 57s | 4m 5s | 2m 52s faster (Kimi K3 256K leads) |
| Approx. task cost | $0.077 / task | $0.138 / task | $0.061 / task lower (Grok 4.7 leads) |
Grok 4.7: 5.0% of SuperGrok Plus (5.0% / 1w) per run. Kimi K3 256K: 115.0% of Kimi Allegretto (115.0% / 5h, 23.0% / 1w) per run. Its weekly limit holds about 5 of the week's 33.6 5h windows. Each figure is a share of that agent's own plan, so there is no delta to draw.
| Dimension | Grok 4.7 | Kimi K3 256K | Advantage |
|---|---|---|---|
| Capability | 68% | 53% | +15% (Grok 4.7 leads) |
| Median task time | 37m 24s | 5m 36s | 31m 49s faster (Kimi K3 256K leads) |
| Approx. task cost | $0.402 / task | $0.247 / task | $0.156 / task lower (Kimi K3 256K leads) |
Grok 4.7: 7.0% of SuperGrok Plus (7.0% / 1w) per run. Kimi K3 256K: 58.0% of Kimi Allegretto (58.0% / 5h, 11.0% / 1w) per run. Its weekly limit holds about 5 of the week's 33.6 5h windows. Each figure is a share of that agent's own plan, so there is no delta to draw.
| Dimension | Grok 4.7 | Kimi K3 256K | Advantage |
|---|---|---|---|
| Capability | 64% | 24% | +40% (Grok 4.7 leads) |
| Median task time | 11m 10s | 6m 33s | 4m 38s faster (Kimi K3 256K leads) |
| Approx. task cost | $0.092 / task | $0.269 / task | $0.177 / task lower (Grok 4.7 leads) |
Grok 4.7: 0.4% of SuperGrok Plus (0.4% / 1w) per good solve. Kimi K3 256K: 14.0% of Kimi Allegretto (14.0% / 5h, 3.0% / 1w) per good solve. Its weekly limit holds about 5 of the week's 33.6 5h windows. Each figure is a share of that agent's own plan, so there is no delta to draw.
Grades and deltas are comparable only within a suite version. See methodology.
Frequently asked questions
Which is better for coding, Grok 4.7 or Kimi K3 256K?
Grok 4.7 clears higher: tier 4 of 5 versus tier 2 for Kimi K3 256K on real-world@v3, our private suite of real-world coding problems. Latest measurement 2026-09-26.
Which is cheaper per task, Grok 4.7 or Kimi K3 256K?
Grok 4.7 used 12% of SuperGrok Plus's weekly limit per pass, $0.145 per task at the listed price. Kimi K3 256K used 34% of Kimi Allegretto's weekly limit per pass, $0.161 per task at the listed price. Per task the two cost about the same ($0.145 vs $0.161).
What do Grok 4.7 and Kimi K3 256K cost a month?
Grok 4.7 runs on SuperGrok Plus at $100/month (public price as of 2026-09-14). Kimi K3 256K runs on Kimi Allegretto at $39/month (public price as of 2026-07-20).